Flow-Profiling for Light Roasts: Controlling Fruitiness Without Losing Body
Flow-Profiling for Light Roasts: Controlling Fruitiness Without Losing Body
Light roasts are denser, less soluble and higher in acidity – and that is exactly why they react more sensitively to what happens in the puck during the first few seconds. Flow-Profiling is an attempt to control that. Here is what the measurements have to say about it, and where practical experience takes over.
Pressure and flow are not independent variables. You only ever control one of the two; the other follows from the permeability of the puck. Depending on the grind size, the same flow setpoint produces anywhere between 2.9 and 8.0 bar.
What is established: a lower flow rate extends the contact time and increases the concentration at the start of the shot. Light roasts really are harder to extract, because of their lower porosity.
What is not established is that Flow-Profiling tastes different at the same TDS and the same extraction yield. There is no peer-reviewed sensory study on this. The case for Flow-Profiling is control and reproducibility – not taste per se. That is our reading, and we think it is the most honest one.
Why pressure and flow are linked
The underlying equation is Darcy’s law: the volumetric flow equals permeability times area times pressure difference, divided by viscosity times puck height. The only variable in it that you do not know directly is the permeability – and it is not a constant.
Corrochano et al. (2015) measured permeabilities from 2.59 × 10−14 to 3.36 × 10−13 square metres. A factor of 17 between the extremes. And a 30 per cent higher bulk density lowered the permeability three- to fourfold.
With a fine grind, the same increase in flow costs more than 5 bar of additional pressure; with a coarse grind, not even 1 bar. If you set the pressure, you hand control of the flow to the puck – and vice versa.
The finding that contradicts Darcy
Peer reviewIn 2026, Waszkiewicz and colleagues measured 60 shots across eleven pressure steps from 1 to 12 bar, accompanied by X-ray microtomography. Their central result is counterintuitive: flow reaches a maximum at around 5 bar of basket pressure and drops again as the pressure keeps rising.
The authors’ explanation: the puck compresses itself under pressure. Higher pressure lowers the permeability so much that it more than offsets the larger pressure gradient. Poroelasticity rather than pure Darcy.
This is the strongest physical argument for Flow-Profiling that we know of: if you set the pressure, you are controlling a variable whose effect on flow is not even monotonic. If you set the flow, you are directly controlling what is tied to contact time. A side finding of the same study: the dissolution dynamics themselves also change the flow during the shot.
What flow changes about extraction
Schmieder et al. (2023) at TU Munich systematically varied flow rate, grind size and temperature and measured trigonelline, caffeine, chlorogenic acid and TDS. Two key findings:
First: of the three factors, flow rate had the strongest effect on the amount of substance in the cup. A higher flow rate reduced the mass of all measured components. The reason is simply contact time – slower flow, longer contact, higher concentration at the start.
Second, and this is the uncomfortable half: the authors themselves write that the effect of flow rate, grind size and temperature is small compared with the influence of the brew ratio. According to this study, Flow-Profiling is a fine-tuning tool, not a main lever.
The order of extraction over the course of the shot was: trigonelline and TDS drop off fastest, then chlorogenic acid, with caffeine the flattest. The authors link this to the polarity of the compounds.
This rule of thumb appears in almost every text on espresso. In the specialist espresso literature, we could not find it confirmed in this form. Schmieder et al. measured three individual substances plus TDS – no melanoidins, no bitter compounds, no organic acids. What is established is the order trigonelline, chlorogenic acid, caffeine. Carrying that over to “acidity versus bitterness” is a conceptual model. Useful as a working model, not as a fact.
What is well established, on the other hand, and particularly relevant for light roasts: Smrke, Eiermann and Yeretzian (ZHAW) found that the concentration of volatile aroma compounds does not rise linearly with extraction yield, and that faster extractions preserve highly volatile aroma compounds better than long ones. Turbo shots of 10 to 20 seconds reached 17 to 18 per cent extraction yield – over 80 per cent of the maximum. If you want to preserve fruitiness, this is your most direct starting point.
Why light roasts are harder to extract
The reason is the pore structure. Lindsey et al. (2024) measured porosity across roast levels: from 13.6 per cent in unroasted green coffee to 45.8 per cent in the darkest batch. Light roasts sit in between – and the authors explicitly attribute the inhibited extraction to the lower porosity.
At the same time, the pH is lower. Grim, Fuller and Rao (2020) measured pH values rising with increasing roast across three roast levels. So a light roast means more acidity combined with lower solubility. Exactly the conflict that Flow-Profiling is meant to address.
One detail we find fascinating: Schenker et al. (2000) compared two roast profiles with practically identical roast loss. The fast-roasted batch had a porosity of 0.528, the slow-roasted one only 0.474, with clearly different densities. It is not only the roast level that determines porosity, but also the roast profile. Two equally light coffees can differ measurably in their resistance to extraction – something we regularly experience in our own roastery.
Practical consequences
| Pressure profiling | Flow-Profiling | |
|---|---|---|
| You control | the pressure | the flow |
| Follows on its own | the flow (from the permeability) | the pressure (from the permeability) |
| With grind drift | flow drifts with it, yield drifts | pressure drifts with it, yield stays more stable |
| Pressure spikes | possible if the grind is too fine | implicitly limited |
And one lever that often gets lost in the discussion, even though it is the best supported by evidence: Cameron et al. (2020) recommend lowering the dose from 20 to 15 grams and grinding coarser for more homogeneous extraction, at the same beverage mass. The result: more even extraction, more reproducible shots and 25 per cent less coffee used. The work was done at 6 bar, not 9. For light roasts, this is the most direct published route to a higher extraction yield without over-extraction.
On pressure itself, by the way, there is an older, good sensory study: Andueza et al. (2002) compared 7, 9 and 11 atmospheres. At 11, the espresso became bitter and astringent with more pronounced off-notes; at 9, the foam consistency and the proportion of fruity and fresh odorants were best; at 7, the foam was unstable. So the classic 9 bar is not arbitrary – at least with a fixed grind size.
According to our research, there is no published evidence on this. And a finding from a related field calls for caution: Batali et al. (2020) found that, with TDS and extraction yield held constant, brew temperature had practically no effect across 31 sensory attributes. If process parameters fade away to that extent at fixed TDS and fixed extraction yield, the question for Flow-Profiling is at the very least open. That is why we frame it as a hypothesis, not as a selling point.
Our practical recommendation – and it is exactly that, a recommendation, not a rule: use Flow-Profiling as a tool for control and reproducibility. Set the brew ratio and dose first, then the grind size, and only after that the profile. You can play through what the profile does in the pressure profiling simulator and the extraction simulator. What happens before the profile we have described under Pre-infusion.
Frequently asked questions
What is the difference between Flow-Profiling and pressure profiling?
With pressure profiling, you set the brew pressure and the flow follows from the permeability of the puck. With Flow-Profiling, you set the flow and the pressure follows. The two are coupled through Darcy’s law – you only ever control one of the two variables.
Does Flow-Profiling bring out more fruitiness in light roasts?
There is no peer-reviewed sensory study on this. What is established is that slower flow extends the contact time and increases the concentration, and that faster extractions preserve highly volatile aroma compounds better. The case for Flow-Profiling is above all control and reproducibility.
Why are light roasts harder to extract?
Because of their lower porosity. Lindsey et al. (2024) measured porosities from 13.6 per cent in green coffee to 45.8 per cent in a dark roast and explicitly attribute the inhibited extraction of light batches to this. On top of that comes the lower pH, meaning more acidity combined with lower solubility.
Is 9 bar really the right brew pressure?
Andueza et al. (2002) compared 7, 9 and 11 atmospheres: at 11, the espresso became bitter and astringent; at 7, the foam was unstable; at 9, foam consistency and fruity aroma compounds were best. More recent work, however, shows that flow reaches its maximum at around 5 bar of basket pressure – more pressure does not mean more flow.
What helps more than profiling?
Going by the research, the brew ratio. Schmieder et al. (2023) themselves write that the influence of flow rate, grind size and temperature is small compared with the brew ratio. A second well-supported lever: reduce the dose and grind coarser at the same beverage mass – this gave more homogeneous extraction with 25 per cent less coffee.
Roastery tip
Light roasts show profile changes most clearly. Wild Peach is our reference coffee for this – with it, you can tell straight away whether the acidity has structure or is just sharp. Nutty Delight works well as a cross-check: if a profile changes nothing there, it is down to the profile and not the coffee. Our recommendation, one of many.
